A Neuropsychological Approach to Intelligence

Transcription

1 Neuropsychology Review. Vol. 9. No A Neuropsychological Approach to Intelligence Alfredo Ardila 1,2 Th~s.paper proposes that current 'psychometric intelligence tests are limited in evaluating cognitive ac~i~~ty. From a neuro~sychol~glcal perspective. they fail to measure some fundamental cognitive ablh~es such as executive functions. memory. and visuospatial abilities. The analysis of the Wechsler Intelhg~nce Scale presented shows that the original rationale for selecting the specific sub tests included ~n the WAIS was unclear. The concept of a g factor in cognition is also analyzed. with the conclusion that the g factor continues to be controversial. The value of intelligence tests in predicting school perfonnance is also criticized. It is proposed that the psychometric concept of general intelligence should be deleted from cognitive and neurological sciences. Finally, it is proposed that, in the future. neuropsychological instruments sensitive to more specific cognitive abilities replace current psychometric intelligence tests. KEY WORDS: Intelligence; IQ; neuropsychological assessment; cognitive abilities., I I I I! i I I! i i I! I l INTRODUCTION Curiously. intelligence tests do not appraise intelligence. (Anonymous) The idea that is presented in this paper is very simple: From a neuropsychological perspective, current psychometric intelligence tests (e.g., Wechsler Adult Intelligence Scale; WAIS) are limited in evaluating cognitive abilities. Furthermore, when using compound scores (e.g. IQ), it is not sufficiently apparent what these general scores measure. It is suggested that in the future, current testing methods be replaced by neuropsychological cognitive assessment instruments. This idea, developed throughout the paper, emerges from the following points. (1) There are two different sets of instruments directed to the appraisal of cognitive abilities: psychometric intelligence tests (e.g., the WAIS in its different versions) and neuropsychological assessment tests (e.g., Luria-Nebraska Neuropsychological Battery, NEUROPSI). No evident reason seems to exist to maintain this duality. It may be argued that psychometric intelligence tests are directed to normal populations, whereas neuropsychological instruments are directed to I Instituto Colombiano de Neuropsicologia, Bogota, Colombia. 2Correspondence regarding this article should be addressed to the author at NW 8 Street. Miami. Florida brain-damaged populations. Neither argument is accurate. Psychometric intelligence tests are frequently included in the neuropsychological evaluations of brain-damaged individuals. Even a Wechsler Intelligence Scale adapted for neuropsychological purposes has been developed (WAIS R-NI; Kaplan et al., 1991). Neuropsychological instruments can be and are frequently used with normal populations. Initially, neuropsychological tests are administered to normal subject populations (norming studies) before being used with abnormal subject populations. It could be further argued that neuropsychological tests are frequently easy and often have a low ceiling. Indeed, neuropsychological tests target pathological people and, in neurologically normal people, the ceiling frequently is rapidly reached. The ceiling effect varies depending on the specific test. However, this is not an intrinsic limitation, and the ceiling can be raised. (2) From a neuropsychological point of view, intelligence tests do not evaluate some abilities that should be included as "fundamental cognitive abilities" (i.e., "intelligence"). An analysis of executive functions (i.e., "frontal lobe" abilities), memory, and visuospatial abilities is presented in this paper. It is emphasized that according to contemporary neuropsychology, these abilities represent some of the most important cognitive abilities. They are inappropriately tested using current psychometric instruments. (3) There was not sufficient scientific rationale for selecting the set of subtests 104G-7308I99f $ 16.OWO <C 1999 Plenum Publishing Corporation

2 118 Ardila included in current intelligence batteries. Most of our current knowledge about the brain organization of cognition has been obtained during the past 50 years. The specific WAIS subtests currently in use were selected prior to acquiring this knowledge-nearly three quarters of a century ago. The analysis of intelligence testing presented in this paper will center on the Wechsler Intelligence Scales (WIS). There are several reasons for this. First, it represents the most widely used intelligence scale, not only in the United States but also in many other countries. Second, it is the best studied and analyzed intelligence scale. Many research studies using the WIS in different areas are easily available. Third, there are different ver- \ions of the test (WAIS-R, WISC-ill, WAIS-ill, etc.) that have attempted to overcome the shortcomings that existed in previous versions. As a result, they can be considered the best designed intelligence scales to date; at the least, very significant amounts of time and effort have been devoted to their de~ign, redesign, and use. And fourth, they are the most frequently used intelligence test battery components in neuropsychology. Researchers have even adapted the WAIS to the neuropsychological perspective (WAIS-R NI; Kaplan et al., 1991). It has to be emphasized that not all the concerns presented with regard to the WIS are applicable to other intelligence test batteries. For example, the second most popular intelligence scale, the Stanford-Binet Intelligence Scale (Thorndike et al., 1986) uses a somewhat different approach to intelligence. Stanford-Binet postulates a three-level hierarchical model of intelligence: The first level is represented by a general intelligence factor (g). There are three second level factors (Crystallized Abilities, Fluid-Analytic Abilities, and Short-Term Memory). Crystallized Abilities include Verbal Reasoning and Quantitative Reasoning. The third level refers to the subtests included in this intelligence scale: Vocabulary, Comprehension, Absurdities, and Verbal Relations to evaluate Verbal Reasoning; Quantitative, Number Series, and Equation Building to assess Quantitative Reasoning; Pattern Analysis, Copying, Matrices, and Paper Folding and Cutting to appraise Fluid-Analytic Abilities (Abstract-Visual Reasoning); and finally Bead Memory, Memory for Sentences, Memory for Digits, and Memory for Objects to evaluate Short-Term Memory. It is evident that the Stanford-Binet overtly recognizes that reasoning and memory represent fundamental elements of cognition. The concept of intelligence can be criticized from two different points of view. (1) Cognitive abilities measured by intelligence psychological tests represent, at least in their contents, culturally learned abilities. Performance is influenced by a vast array of moderating variables, including culture, ecological demands, primary language, and educational level. Test scores are associated, therefore, not only with the subject's learning opportunities, but also with those variables that a culture dictates worthy of cognitive amplification (Ardila, 1995a). Different cultural environmental contexts will result in the development of different patterns of abilities (Berry, 1971, 1979). Further, when tests are used with members of a different culture, testees often do not share the presumptions about values, knowledge, and communication implicitly assumed by the test (Greenfield, 1997). (2) The specific tasks used to tap intelligence are inappropriate. The first point has been extensively analyzed, particularly in anthropology (e.g., Irvine and Berry, 1988) and cross-cultural psychology (e.g., Berry et al., 1992). In this paper, the major emphasis will be placed on the second point (i.e., that the specific tasks used to tap intelligence are inappropriate). A general conclusion of this paper is that the concept of general intelligence should be abandoned. But abandoning the concept of intelligence may seem too extreme. Intelligence has become a fundamental cornerstone of contemporary psychology. Of course, there are two different issues in this regard: the word intelligence and the existence of a general factor (g) in cognition. It is proposed that the word intelligence be replaced with either cognitive abilities or simply cognition. What is the difference? The answer is simple: intelligence is confusing and difficult to operationalize. Furthermore, it has been frequently equated with psychometric intelligence tests (IQ). It is also pointed out that the existence of a g factor in cognitive testing is questionable. An examination of the history of the g factor in intelligence, and the different factor-analytic studies carried out during the last decades, leads to the conclusion that the assumption of a g factor is difficult to sustain. Many researchers consider that g-based factor hierarchy confusing and misleading (e.g., Ceci, 1990; Lezak, 1995). Finally, it has to be emphasized that intelligence is obviously a construct, not a physical entity. Frequently, however, the term intelligence has been used as if it were a physical entity, a reification that can be easily and objectively measured. In this paper, it is argued that the construct of intelligence is no longer tenable, and thus, measures of general intelligence are inappropriate and misleading. Many words from popular language have been incorporated into psychology. Terms such as will, mind, and consciousness are just a few examples that have been eliminated with the evolution of psychology because their precise meaning is vague. For the same reason, it is proposed that the term intelligence should also be removed. br

3 Intelligence and Neuropsychology 119 INTELLIGENCE TESTING: A BRIEF HISTORICAL OVERVIEW The attempt to measure intelligence represents one of the major endeavors in 20th century psychology. A tremendous amount of research has been directed to unders~ding the organization of intellectual activity and discussing the procedures appropriate to its measurement. In 1904 the Ministry of Education in France commissioned Alfred Binet and Theophile Simon to develop a practical procedure to distinguish between mentally retarded and normal children at school. To fulfill this purpose, they developed a kind of developmental scale describing the types of abilities that were normally expected at di{ferent ages (Binet, 1905,1908). The concept of "mental age" was introduced to refer to the level of development expected at each age. Later, Stern (1912) introduced the concept ofiq. The Binet-Simon tests were rapidly adopted in England, the United States, and other countries. In the United States, Terman (1916), at the University of Standford, adapted and standardized the scales presented by Binet and named them the Stanford Revision of the Binet Scale, or simply Stanford-Binet. Terman also further developed the concept ofiq. Nonetheless, how to understand cognitive abilities measured in intellectual tests remained significantly controversial. Two different interpretations of cognitive abilities rapidly became evident: (1) There is a general intelligence factor that potentially may be measured and even quantified; and (2) there are different cognitive abilities, not a single one. That is, single compound scores are not acceptable. The "g" Intelligence Factor Spearman (1904, 1923) may be considered the most important representative of the first point of view. He hypothesized a two-factor theory of intelligence. He supposed that any test measures a g factor common to all other cognitive tests; and a specific factor (s) unique to that particular test. The relation between g and s components may be variable, but g is always included in any cognitive tests. Tests without the g factor may be tests of sensory or motor abilities, but they do not represent cognitive tests. The existence of this g factor constitutes the theoretical basis to accept that intelligence can be quantitatively measured using a simple score (lq). Spearman's theory was subjected to diverse fundamental criticism on empirical grounds. However, "while Spearman was aware that his theory had been empirically refuted, he continued to emphasize the importance of a common factor in intelligence" (Brody, 1992, p. 13). Multiple-Factor Approaches The most distinguished representant of the second point of view was L. L. Thurnstone (1938, 1947), who further developed factor analysis, attempting to obtain the most parsimonious solutions of the data to which they were applied. He introduced new concepts and more sophisticated procedures in factor analysis, such as oblique-factor structure and centroid methods. He proposed a relatively limited number offactors that would correspond to the fundamental or primary mental abilities: Space, Verbal Comprehension, Word Fluency, Induction, Perceptual Speed, Deduction, Rote Learning, and Reasoning. He supposed that each factor should correspond to certain specific nervous system activity. Further studies (e.g., Kaiser, 1960) have significantly supported most of the original primary factors proposed by Thurstone. Vernon (1950) developed a kind of hierarchical model to describe the organization of cognitive abilities. He assumed two major or second-order factors to group primary abilities or primary factors: Verbal-Educational (v:ed) and Spatial-Mechanical (k:m) abilities. Thus, there was a kind of hierarchy in intelligence from the most general factor g, to the major factors v:ed and k:m, to the minor factors or primary abilities, and finally to those factors specific to each test. Guilford (1967,1968; Guilford and Hoepfner, 1971) took a somewhat different approach. He proposed a three-dimensional classification of intelligence including contents (letters, numbers, words, and behavioral descriptions); operations (memory, evaluation, convergent thinking, and divergent thinking); and products (units, classes, relations, systems, transformations, and implications). Consequently, according to Guildford, 120 different intellectual abilities could be distinguished. He supposed that empirical data would support the existence of this high number of intellectual abilities. Cattell (1971) proposed that more than one secondorder analysis factor could be found. He distinguished between "Fluid Intelligence" (corresponding to and reflecting a pattern of neurophysiologica.l and incidental learning influences) and "Crystalized Intelligence" (highly sensitive to each person's unique cultural, educational, and environmental experiences). This distinction rapidly became quite popular. Cattell's distinction between two different types of intelligence is similar to the two major intellectual factors proposed by Hebb (1942): Intelligence A and t

4 120 Intelligence B. Intelligence A represents the basic biological ability to acquire knowledge. Intelligence B reflects the influence or expression of acculturation, education, and personal experiences. The Idea of Multiple Intelligences Gardner's multiple intelligence approach might be interpreted as a return to Thurnstone. Gardner (1983) proposed the existence of different and independent types of intelligence. In developing his model of intelligence, he began with several observations: (1) Damage in different neural structures may result in impairing certain abilities while sparing other abilities; that is, isolated defects it1 some types of cognition can be observed in cases of brain pathology (a procedure known in neuropsychology as "double dissociation"). (2) Individuals such as idiotsavants demonstrate a significant dissociation in different cognitive abilities. That is, in non-brain-damaged individuals intellectual abilities may be dissociated, and even extremely dissociated. (3) Every type of ability is identified by a specific set of operations related to a neural mechanism; in this regard, Gardner is attempting to reconcile the idea of several types of intelligence with current research about brain organization of cognition. (4) A specific developmental history for each type of intelligence; that is, different cognitive abilities ("intelligences") develop independently in a child. (5) An evolutionary history exists for each intelligence; that is, different intelligences may have different origins in subhuman species and may have evolved in different ways. (6) Experimental psychology supports the existence of different intelligences. (7) Psychometric studies support the independence of different cognitive abilities; Gardner insists that psychometric research has not investigated widely enough the diversity of intellectual abilities that are observed in real contexts. (8) Susceptibility of different abilities to encoding in a symbolic system; he proposed that cognitive abilities tend to be encoded in culturally different devised symbolic systems. Departing from these considerations, Gardner proposes six different types of intelligence: Linguistic, musical, logic-mathematical, spatial, body-kinesthetic, and personal. This group of intelligences may partially correspond to Thurnstone's primary mental abilities. However, Gardner is relying not simply on psychometric procedures but also on a broad array of contemporary research, including contemporary neuropsychology. In brief, Gardner proposes a relatively limited number of basic and independent abilities or types of intelligence. The similarity with Thurnstone's primary mental Ardila abilities, and contemporary factor analytic studies of neuropsychological tests, is remarkable. Sternberg's Triarchic Theory Sternberg (1988) defined intelligence as "the mental activity underlying purposive adaptation to, shaping of, and selection of real-world environments relevant to one's life" (p. 69). The triarchic theory of intelligence consists of three closely interrelated subtheories: a contextual subtheory, a componential subtheory, and an experiential subtheory (Sternberg, 1985). The contextual subtheory limits intelligence to mental activity underlying environments relevant to one's life. In consequence, intelligence should be conceptualized considering the real conditions existing in the immediate environment. Intelligence. represents adaptation to one's environment. Mental activity is directly inferable through techniques widely available to cognitive psychology. The componential subtheory states that the mental mechanisms are those that affect and are affected by context. Intelligence makes sense only within a particular context. The basic "mental unit" of analysis in this subtheory is the information processing component. This refers to the process transforming sensory inputs into conceptual representations, transforming a conceptual representations into another, or transforming conceptual representations into motor acts. The experiential subtheory states that tasks are particularly relevant to the measurement of intelligence when they measure cognitive performance either when a task or situation is novel or when the task is in the process of becoming automatized. Sternberg (1997) has attempted to apply his interpretation of intelligence to testing in the field of intelligence and the understanding of lifelong learning. His interpretation of intelligence allows significant cultural variations and emphasizes the understanding of the behavioral context. A Processing Speed Interpretation of Intelligence It has also been proposed that intelligence depends on what may be called "the neural efficiency of the brain" (Eysenck, 1986). Several recent studies have demonstrated that the time required to perform some simple perceptual tests are significantly correlated with psychometric intelligence test scores. This means that intelligence may be related to some characteristics of information processing in the central nervous system. Jensen (1987) observed a

5 I~telligence and Neuropsychology correlation between choice reaction time and scores on intelligence tests. These correlations, however, were not particularly impressive (about to -0.30). It was observed that reaction time was inversely correlated with IQ and measures thought to singly predict approximately 10-15% of the variance in IQ (Brody, 1992). Higher correlations on the order of using more complex reaction 6me techniques have been reported by Frearson and Eysenck (1986). Nettlebeck (1987) found a correlation on the order of between inspection time and IQ. The technique used consisted of tachistoscopic presentations of two adjacent vertical lines followed by a masking stimulus. The time of exposure varied and a psychophysical function was obtained... The task was used to ascertain the minimal expo sure time required to obtain a certain level of accuracy in recognizing which one of the two lines was longer. Reed and Jensen (1992) have used visual evokedpotentials to assess what they call nerve conduction velocity. They calculate this velocity by dividing the subject's head length by the latency of an early visual evoked potential component. Using this procedure, they report a correlation between nerve conduction velocity and intelligence on the order of 0.20 to In brief. some significant correlations have been established between speed in information processing and scores on psychometric intelligence tests. These measures, however, usually predict only a relatively modest percentage of the variance. Neuropsychologically Oriented Intelligence Tests Some attempts have been made to approach the concept of intelligence and to develop intelligence test batteries based on a neuropsychological perspective. Two of these attempts will be briefly examined: The Kaufman Adolescent and Adult Intelligence Test (KAIT; Kaufman and Kaufman. 1993, 1997) and the Cognitive Assessment System (CAS; Das et ai., 1994; Naglieri and Das. 1996). The KAIT provides three types of scores: Auid. Crystalized. and Composite IQS. It is applicable to people between the ages of 11 and 85. According to the authors, the tests were developed based on the models of Piaget's formal operations and Luria's planning ability in an attempt to include high-level decision making tasks (Luria's third functional unit). The Crystalized Scale includes Definition. Auditory Comprehension, Double Meaning, and Famous Faces subtests. The Auid Scale includes Rebus Learning. Logical Steps. Mystery Cards. and Memory for Block Designs. The KAIT also include two additional subtests (Rebus Delayed Recall and Auditory Delayed Recall) 121 and a supplement test (Mental Status). Each IQ (Auid, Crystalized, and Composite) has a mean of 100 and standard deviation of 15. Naglieri and Das (1996) suggested that intelligence should be seen as a cognitive construct, integrating neurophysiological findings, cognitive processing research, and sociocultural components of human performance. They base their intelligence theory on Luria's interpretation about the three brain functional units (motivationemotion, processing-storing information, and planningcontrolling behavior). They assume that intelligence consists of these three components: attentional processes that provide focused cognitive activity, information processes of two types (simultaneous and successive), and planning processes that provide control of attention; the use of information processes, internal and external knowledge, and cognitive tools; and self-regulation to achieve desired goals (Naglieri, 1997). They refer to their theory as the Planning, Attention, Successive, Simultaneous (PASS) theory of intelligence (Das et al., 1994). They then developed a Cognitive Assessment System (CAS) applicable to children up to the age of 18. The CAS includes measures of attention (Expressive Attention, Number Detection, Receptive Attention), simultaneous processing (Matrices, Figure Memory, Verbal-Spatial Relations), successive processing (Word Series, Sentence Repetition, Sentence Question, Speech Rate), and planning (Number Matching, Planned Codes, Planned Connection). Both test batteries have at least three major common points: (1) They relate intelligence with brain activity and in this regard represent neuropsychologically oriented intelligence scales; (2) they are based on Luria's theory about brain organization of cognition; and (3) they attempt to include those cognitive abilities associated with prefrontal functions (Luria's third functional unit; prefrontal or "executive" functions). In this regard, they recognize that executive functions must be regarded as crucial elements of intelligent behavior. THE WECHSLER INTELLIGENCE SCALES What is intelligence? Many definitions of intelligence have been proposed (e.g., Binet, 1908;' Jensen, 1980; Sternberg, 1985; Wechsler, 1944). In current literature, we still find a wide variety of definitions, many of which make reference to the mental abilities. For the purpose of this analysis of intelligence tests, Wechsler's definition will be used. Wechsler (1944) defined intelligence as "the aggregate or global capacity of the individual to act purposefully, to think rationally and to deal effectively with his environment" (p. 3).

6 122 This definition can be divided into four different elements: (I) Intelligence is an aggregate or global capacity, (2) to act purposefully, (3) to think rationally, and (4) to deal effectively with the environment. The first element of Wechsler's definition of intelligence refers to a core issue: Is there such a thing as a global or general intelligence, or rather, is intelligence an aggregate of abilities? In h~s defintion, Wechsler does not take a definite position, but he assumes a theory of general intelligence in developing his test battery (Full Scale IQ). Yet he recognizes at least two major types of intelligence: verbal and performance intelligence. This question about one or several intelligences continues up to the present day (see Neisser et 01., 1996). The second element in Wechsler's definition of intelligence ("to act purposefully") could be understood as the control, organization, and planning of behavior. Acting purposefully is evidently a frontal lobe function (executive function) if taken from a neuropsychological perspective (e.g., Luria, 1980; Stuss and Benson, 1986). The third element ("to think rationally") might be understood as either organization of cognition (metacognition) or problem-solving ability. In either case, the definition deals with executive functions (Stuss and Benson, 1986). It means that intelligence, to a significant degree, refers to executive functions. From the neuropsychological perspective, this approach may sound quite attractive (Le., intelligence means planning behavior and organizing cognition). Unfortunately, it will be explained that WIS blatantly fails to evaluate executive functions. There is an overt discrepancy between the defintion of intelligence presented by Wechsler and the testing included in the WIS. The final element ("to deal effectively with the environment") refers to the functional criteria of intelligence. Of course, intelligence may be understood not only from a psychometric perspective but also from a functional perspective (pirozzolo, 1985). Wechsler appropriately recognizes that intelligence has to be considered with regard to the specific environment. The physical and social environment can be quite different between Seattle and the Amazonian jungle, as well as between different subcultures existing in a complex city such as New York. Colombian street children can deal extremely well with their city environment, even though from a psychometric point of view they may score at the level of mental retardation. Unfortunately, it is not so easy to evaluate effectiveness in dealing with the environment from the outside. This can only be appropriately evaluated from inside the culture or subculture itself. In this regard, intelligence becomes an anthropological issue. The question at this point is how Wechsler decided that the WIS subtests he selected were the most appropriate Ardila to evaluate the capacity of the individual to act purposefully, to think rationally, and to deal effectively with his environment. Wechsler (1944) explains that: In arriving at our final choice of tests we used the following procedure: (I) Careful analysis was made of the various standardized tests of intelligence now in use. These were studies with special attention to the author's comments with reference to the type of functions measured. the character of the population on which the scales were originally standardized, and the evidence of the test's reliability. (2) An attempt was made to evaluate each test's claim to validity as evidenced by its degree of correlation (a) with other recognized test and, (b) more importantly still, with subjective ratings of intelligence. The later included teachers' estimates, rating by army officers (as in the case of the Army Alpha and Beta), and estimates of business executives (as in the case of various tests which had been tried out in industry). (3) An attempt was made to rate the tests on the basis both of our own clinical experience and of that of others. (4) Some two years were devoted to the preliminary experimental work of ttying out various likely tests to on the several groups of known intelligence level. (p. 76) Unfortunately, Wechsler fails to explain exactly how these steps were taken. DO INTELLIGENCE TESTS PREDICT SCHOOL PERFORMANCE? The most frequent argument of the defenders of intelligence testing and IQ is that intelligence tests can in a reliable way predict school performance (e.g., Jensen, 1980). As a matter of fact, this was the initial purpose of intelligence tests. The influence of educational variables on intelligence test performance represents a well established observation (e.g., Anastasi, 1988; Cronbach, 1990). Educational attainment significantly correlates with scores on standard tests of intelligence. This correlation ranges from about 0.57 to 0.75 (Matarazzo, 1972). Correlations with verbal intelligence subtests are usually higher (from about 0.66 to 0.75) than correlations with performance intelligence subtests (from about 0.57 to 0.61). But correlation does not mean causality; it simply means association. The crucial question is: Do intelligence tests really predict school performance? Or, do schools train those abilities appraised in intelligence tests? To answer these questions is not easy, even though frequently the interpretation has been that IQ does predict school performance (e.g., Hunter, 1986). Other researchers, however, consider that IQ scores are, to a significant extent, a measure of direct and indirect school learning (e.g., Ceci, 1990). Ceci (1991) presented an extensive and detailed review of available data in this area. The general conclusion is that school attendance accounts not only for a substantial

7 Intelligence and Neuropsychology portion of variance in children's IQ but also apparently some, though not all, of the cognitive processes that underpin successful performance in IQ tests. The magnitude of this influence ranges between 0.25 to 6 IQ points per year of school. As a result, the association between IQ and education cannot be interpreted as indicating that IQ predicts school success. Intelligence and schooling have complex bidirectional relationships, with each one influencing variations in the other (Ceci and Williams, 1997). There are two additional observations that should be emphasized. (1) The largest correlations between IQ and school performance are not found with Full Scale IQ but with Verbal IQ, and particularly with some verbal sub tests (e.-g., Vocabulary). So, the question that might be raised is, Why bother to administer the complete intelligence scale if the verbal subscale (and even one single subtest) is sufficient, and may even be better? And, (2) IQ may "predict" performance in language, reading, writing, and arithmetic but cannot predict performance in other areas, such as drawing or music. Simply speaking, if language tests are used as predictors, verbal performance can be predicted. That is quite obvious. Evidently, our current educational system is significantly biased in favor of verbal abilities. This is not true in other cultures' educational systems, nor is it always true in all of our educational programs. It therefore seems questionable that the WIS is a good predictor of school performance in music and art schools. Ceci (1991) emphasizes that there is a circularity in the association between school performance and IQ. IQ test items were initially selected from those items that according to teachers' opinions, poor learners had the most difficulty answering in class. Variants of these questions are still found in large number on contemporary IQ tests. So, it is obvious that IQ tests would predict school success, as they were composed of items that poor learners found most difficult. Consequently, school failure is both explained as a lack of intelligence and is itself the basis for the definition of lack of intelligence. The circularity is evident. A NEUROPSYCHOLOGICAL PERSPECTIVE OF COGNITIVE ABILITIES AND THEIR MEASURE In neuropsychology a significant opposition to the use of compound IQ scores is frequently observed. Lezak (1995), for example, stressed that neuropsychological observations demonstrate that there are independent intellectual functions; Brain damage can impair certain functions while sparing others. In consequence, compound, global, or total scores can be artificial and meaningless. 123 From a neuropsychological perspective, this represents.an extremely important observation when constructing any theory about the organization of cognitive abilities; independent cognitive abilities can be independently impaired and can independently deteriorate during nermal and abnormal aging. According to Lezak (1995), the concept of intelligence has limited application. The concept of IQ, she notes, represents so many kinds of more or less confounded functions as to be conceptually meaningless. She concludes that the "IQ as a score is inherently meaningless and not infrequently misleading... IQ as a catchword has outlived whatever usefulness it may once have had and should be discharged" (p. 25). Heterogeneity of cognitive abilities is supported by empirical neuropsychological data. This is true not only in abnormal but also normal populations. As an illustration, Ardila et at. (1998) selected a homogenous sample of normal subjects (300-subject sample, aged years; all of them right-handed, middle-class male university students). An extensive neuropsychological test battery was administered including language, memory, perceptual abilities, concept formation, and praxis abilities, Fortyone different scores were calculated. Table I presents the dispersion in scores observed on some of the most "classical" psychological and neuropsychological tests (WAIS, Wechsler Memory Scale, etc.) that were included in this research. It is evident that a particularly high dispersion in scores is found in this completely normal and homogenous population. The ratio between the lowest and highest scores in most tests was around 1 :5-10. In some test scores it was even higher. In the WAIS subtests, dispersion was particularly high in Information, Arithmetic, Block design, Object assembly, and Digit-symbol. It was relatively lower in Digits, Picture completion, and Picture arrangement. Even in young, normal, and highly educated individuals there is a very significant dispersion in the performance of usual psychological and neuropsychological test scores. Significant intersubject and intrasubject differences have to be taken into consideration in any theory about organization of cognition, and in the evaluation of intellectual abilities. Luria's Interpretation of Brain Organization of Cognition No doubt, one of the most influential theoreticians in contemporary neuropsychology has been A. R. Luria. Luria (1980) proposed that cognitive abilities represent "functional systems." The concept of the functional system is understood as a group of interconnected biological operations that produces a particular biological effect. The

8 124 Ardila Table I. Perfonnance of 300 Normal Subjects in "Classical" Psychological and Neuropsychological Tests Test Mean SD Range Ratio" WAIS Infonnation :6.7 Similarities :4.3 Arithmetic :6.0 Vocabulary :4.2 Comprehension :4.5 Digits :3.4 Picture completion :2.6 Picture arrangement :3.6 Block design Object assembly :22.0 Digit-symbol :8.2 WMS Jnfonnation :2.0 Orientation :1.2 Mental Control Logical Memory :4.2 Visual Reproduction :3.1 Associative Learning :8.6 Verbal Fluency Phonologic :5.6 Semantic :3.6 The Rey-Osterrieth Complex Figure Copy : 1.4 Immediate memory :4.0 Finger Tapping Test Right hand :3.8 Left hand :3.7 Reading speed (words/minute) :6.6 WCST Categories achieved Perseverative errors Nonperseverative errors a Ratio between the highest and lowest scores. When the lowest score is zero, the ratio cannot be calculated. From the mathematical point of view, it would be infinite. These conditions are indicated by a dash. functional system is based on a complex dynamic constellation of stages, situated at different levels of the nervous system, which in perfonning an adaptative task, may be changed without the task itself being changed. To write, for instance, represents a complex psychological process (functional system) that requires the participation of multiple areas of the brain; each of these areas makes its particular contribution to the whole system. A focal lesion of the brain will disrupt the ability to write at a particular level (the ability to perfonn the skilled movements required for writing, the spatial organization of writing, the selection of words, the ability to sequence graphemes, etc.). However, such particular focal damage will also disrupt all the functional systems for which that particular operation is required. For instance, the patient will not only have difficulties for the spatial organization of writing but also for the spatial organization of numbers, figures, drawings, etc. In all the functional systems in which the paricular ability is included, the defect will be apparent. The brain damage produces not the loss of a specific cognitive process (functional system), but its disturbance at a specific level. This implies that neuropsychological assessment will be aimed at disclosing the fundamental defects underlying the apparent deficits. For this purpose, it will be necessary to administer to the patients different types of tasks and to analyze how the particular difficulties in performing each one of them are manifested. Clinical-anatomical correlations were widely developed by Luria. As a matter of fact, he is a precursor of the method of the superimposition of lesions to disclose critical areas in a particular type of disorder. His study of 800 patients to determine the critical brain area for phonemic discrimination deficits has become classic. This procedure of superimposing lesions to highlight critical areas responsible for clinical syndromes is extensively used in the present-day neuropsychological research (e.g., Damasio and Damasio, 1989; Kertesz, 1983). Luria strived to establish correlations between brain pathology and disturbances at specific levels of infonnation processing (e.g., phonemic discrimination), not to correlate brain pathology with perfonnance in specific tests. Tests may be changed, but since some specific level of information processing would still be required, impairment will be manifested. Because performance on even apparently very simple tests can require the participation of different brain systems, perfonnance on such simple tests can be altered as a consequence of very different brain pathology, although the specific errors will be different. Many different types of brain pathology can alter, for instance, calculation abilities; however, in each case the difficulty (and the errors) will be the result of a disturbance at a different level. Patients with frontal lobe damage and patients with angular gyrus damage can both present with serious difficulties in performing simple calculation tests. However, the underlying impaired mechanism and the type of errors manifested are quite different (Rosselli and Ardila, 1989). Consequently, the validity derived from correlating the site of the brain pathology with perfonnance on a particular test appears, in Luria's interpretation, as a very crude approximation. For Luria, the information collected from the observation of brain-damaged patients should be helpful for developing a more accurate picture of brain organization of cognitive processes. If we knew well enough how the brain works we should be able to accurately predict brain pathology when analyzing in detail the perfonnance of a patient on a set of tests. The departure point in the neuropsychological assessment is the knowledge about how

9 Intelligence and Neuropsychology 125 Table U. Factors Underlying Different Aphasia Syndromes, According to Luria (1976a)" Aphasia Type Acoustic-Agnostic Acoustic-Amnesic Amnesic Semantic Afferent Motor Efferent Motor Dynamic Impaired Factor Phoneme discrimination Verbal memory Semantic structure of words Understanding logical-grammatical (quasi-spatial) structures Articuleme discrimination Disturbances in speech kinetic structure Verbal initiative a ''Factor'' in Luria's theory refers to the fundamental defect responsible for a particular neuropsychological syndrome. the;, brain works, not the knowledge about how to apply a series of tests in standardized conditions. It is interesting to note that Luria extensively, but not systematically, used the term factor to refer to the deficit that can underlie an overt clinical disorder. At other times he simply referred to the basic deficit or underlying defect affecting normal psychological performance. And, undoubtably, he did not use the term factor with a mathematical meaning (i.e., factor analysis). At this point, the question arises as to which "factors" underlay performance in different neuropsychological tests. These factors would, in consequence, represent the basic elements of cognition. Luria discussed this question in some detail with regard to language. In his last book Basic Problems of Neurolinguistics" (1976a), Luria analyzed the factors that can underlie the different aphasic syndromes (Table II). As a matter of fact, these same factors had been previously pointed out by Luria years before. However, it is not easy to deduce the impaired factors in other neuropsychological syndromes (e.g., agnostic or apraxic disorders). This "factorial theory" of cognitive activity represents one of the most interesting and outstanding points in Luria's neuropsychological perspective. Unfortunately, Luria did not completely develop this factorial theory of psychological activity. Further Developments Luria's interpretation of brain organization of cognitive activity has received support from contemporary researchers in the area. Thus, for Benson (1994), any complex psychological activity requires the participation of different brain areas. As an example, according to Benson (1982), six different brain areas participate under normal conditions in reading aloud: (1) primary visual cortex, (2) association visual cortex, (3) angular gyrus, (4) temporal areas involved in language recognition, (5) the frontal language area (Broca's area), and (6) the primary motor cortex controlling language articulation. Benson points out that depending on the material used in reading, other additional brain areas could also be involved in the reading process. This whole array of brain areas would represent the "brain system" (Ardila, 1995b) underlying the reading aloud process, and supporting the "functional system" for reading, according to a Lurian interpretation. In case of damage in any of these written language processing levels, a deficit in reading will appear, even though it would be different depending on the specific impaired area. Furthermore, other abilities also relying on one of these processing levels (''factors'') would be also affected. Contemporary neuroimaging and electrophysiological techniques have provided most valuable information about brain activity during performance of different cognitive tasks. Thus, departing from measures of focal brain metabolism, positron emission tomography allows one to visualize levels of brain activity and focal involvement during different conditions. It has been observed that when performing complex intellectual tasks (e.g., reading aloud, speaking, etc.) a complex matrix of activated areas is revealed (posner et al., 1988). Different brain areas participate, making specific contributions to the performance during, for example, a reading task: occipital (visual perception), temporal (language decoding), and frontal Broca's area (language control and production) (Pettersen et al., 1989). For each one, a somehow limited region is fully activated, where some other areas can be only partially active (Price et al., 1994). While speaking, a specific activation of the left mouth area can be observed, as well as in the superior temporal lobe and the supplementary motor area. Factor Analysis in Neuropsychology In the neuropsychological domain, factor analysis has been more frequently applied to some specific tests and scales directed to measure single cognitive abilities. For example, several factor analytic studies of the Wechsler Memory Scale are available to date (Ardila and Rosselli, 1994; Bornstein and Chelune, 1988; Elwood, 1991; Roid et al., 1988; Wechsler, 1987). Nonetheless, factor analyses of extensive neuropsychological battery tests are scarce. Pont6n et al. (1994) administered a neuropsychological test battery including 10 different tests to 300 normal subjects. A factor analysis was used and five different factors were found: a Verbal Factor (measured basically through verbal fluency and naming), a Learning Factor (measured specially with an auditory verbal learning test), a factor related to the Speed in Processing Information

10 126 Ardila (attention; measured with Digit-symbol subtest), a Visual Processing Factor (measured with the Rey-Osterrieth Complex Figure), and finally, a Psychomotor Speed Factor (measured with the Pin Test). Ardila et al. (1994) administered a general neuropsychological test battery to a 98-subject sample. Their battery included language, memory, spatial abilities, concept formation, and praxis abilities tests. A factor analysis with varimax rotation found nine different factors accounting for about 70% of the variance. Factor I (Verbal Factor; accounting for 14.2% of the variance) was measured by a Sequential Verbal Memory Test and Verbal Fluency subtests. Factor IT (accounting for 12.9% of the variance) was measured by the WMS Visual Reproduction subtests (Nonverbal Memory and Constructional Factor; imme- ~iate and delayed reproduction) and the Rey-Osterrieth Complex Figure (copy and immediate reproduction). Factor III (Verbal Memory Factor; accounting for 9.8% of the variance) was measured by the WMS Logical Memory subtests (immediate and delayed). Factor IV (Fine Movements Factor; accounting for 6.4% of the variance) was associated with fine movements (tapping subtests, right and left hand). Factor V (Verbal Knowledge; accounting for 6.0% of the variance) was mainly measured by the Information subtest of the WMS and the Boston Naming Test. Factor VI (Praxis Ability Factor; accounting for 5.5% of the variance) represented ideomotor praxis tests. Factor VII (Delay Associative Learning Factor; accounting for 5.4% of the variance) was measured by the Delayed Associative Learning subtest, and Factor VIII (Arithmetic Factor; accounting for 5.0% of the variance) was measured by Digit Span. Factor IX (Mental Control Factor; accounting for 4.4% of the variance) was measured by the Mental Control subtest of the WMS. Correlations between some tests were negative (e.g., between Logical Memory from the WMS and Rey-Osterrieth Complex Figure-Copy condition). Several correlations were around zero. This observation is particularly important from the point of view of the existence of a general factor in cognition (g factor). Ardila et al. (1998) administered a comprehensive neuropsychological test battery: language, memory, perceptual abilities, concept formation, and praxis abilities tests to 300 normal subjects. Forty-one different scores were calculated. It was found that some of the tests presented a complex intercorrelation system, whereas other tests presented few or no significant correlations. A factor analysis with varimax rotation of the neuropsychological battery tests was performed. Five different factors accounted for 63.6% of the total variance. Factor I (26.7% of the variance) represented a clearly Verbal factor. Factor II was a perceptual or Nonverbal factor (12.5% of the variance). Factor III (9.8% of the variance) correlated with the different scores of the Wisconsin Card Sorting Test. Factor IV (7.9% of the variance) was a Fine Movements factor, and Factor V (6.7% of the variance) represented a Verbal Memory factor. Ostrosky et al. (1999) administered a short neuropsychological test battery assessing a wide spectrum of cognitive functions including orientation, attention, memory, language, visuoperceptual abilities, and executive functions. Normative data in an 800-subject sample from 16 to 85 years of age, and from zero to 24 years of education were obtained. A factor analysis with varimax rotation disclosed seven different factors. Factor I (accounting for 28.6% of the variance scores) best correlated with Digits backwards, Copy of a semicomplex figure, Calculation abilities, and Language Comprehension. Factor II (9.6% of the variance) highly correlated with the writing scores. Factor III (accounting for 6.1 % of the variance) best correlated with verbal fluency tests. Factor IV (accounting for 5.7% of the variance) was correlated with motor functions. Factor V (accounting for 4.3% of the variance) correlated with all the recall scores. Factor VI (accounting for 3.9% of the variance) correlated with Orientation in Space. Factor VII (accounting for 3.6% of the variance) correlated with Orientation in Person. In summary, several factor analyses of extensive neuropsychological test batteries have yielded quite similar results: Some 5-10 factors are found, accounting for about two-thirds of the total variance. The first factor accounts for some 15-30% of the total variance, and it is usually a verbal factor. Some additional factors are also observed: spatial, memory, perceptual, fluency, motor skills, etc. The damage in the "brain systems" supporting the intellectual activities corresponding to these factors (verbal abilities, spatial abilities, verbal fluency, etc.) would result in specific neuropsychological syndromes (aphasia, spatial agnosia, amnesia, etc.). These factors should be matchable with the neuropsychological syndromes found in cases of brain pathology (see Table III). Nonetheless, it does not seem realistic to suppose that the exact number of these factors can be determined. From a factor analysis perspective, the factors to be found obviously depend not only on the types oftests that are included but also on some additional variables (e.g., the subjects included in the analysis, the type of factor analysis, etc.). From a neuropsychological perspective, to exactly pinpoint the cognitive syndromes observed in cases of brain pathology is not an easy task, at least at the moment. As an illustration of this point, we do not know well enough how to classify spatial disturbances associated with brain pathology (e.g., Benton, 1989; De Renzi, 1982, 1985; Hecaen, 1962; Hecaen and Albert, 1978; Morrow and Ratcliff, 1988; Newcombe and Ratcliff, 1989). We do not know yet well enough either

11 Intelligence and Neuropsychology 127 Table m. Factors Observed in Ardila et a1. (1994) Neuropsychological Test Battery. the Tests Most Saturated by These Factors. and Probable Neuropsychological Syndromes.that Might Be Associated with Impainnents in Those Factors Factor Test Probable Neuropsychological Syndrome I Verbal Production Verbal fluency Convexitalleft prefrontal syndrome II. Constructional-Vtsuospatial Rey-Osterrieth Figure Constructional apraxia Visual memory WMS Spatial agnosia III Verbal Memory Logical memory Verbal amnesia Wernicke aphasia IV Fine Movements Tapping test Premotor syndrome Kinetic apraxia V Verbal Knowledge Infonnation Wernicke aphasia Boston Naming Test Anomia VI Praxis Ability Ideomotor apraxia test Ideomotor apraxia VII Delayed Associative Learning Delayed Associative Hippocampal amnesia Learning WMS VIII Arithmetic Digits Acalculia IX Attentional Mental control WMS Orbital prefrontal syndrome Table IV. Some Relatively Constant Factors Found Across Different Factor-Analytic Studies (Carroll. 1993) and the Neuropsychological Syndromes with Which They Might Be Associated Factor Language Lexical Knowledge Grarrunatical Sensitivity Communication Ability Oral Production Speech Sound Discrirnination Naming Facility Expressional and Word Fluency Reasoning Sequential Inductive Quantitative Visual Perception Spatial Relations Serial Perceptual Integration Perceptual Speed Numerical Number Facility Attention and Concentration Attention and Concentration Neuropsychological Syndrome Wernicke aphasia Broca aphasia Prefrontal syndrome Verbal apraxia? Word deafness Anomia Extransylvian motor aphasia Prefrontal syndrome Prefrontal syndrome Frontal acalculia Spatial agnosia Topographic agnosia? Visual agnosia Acalculia Prefrontal syndrome how exactly to classify language disturbances associated with brain damage (e.g., Benson and Ardila, 1996). And we do not know well enough the exact variants of the prefrontal syndrome (Damasio and Anderson, 1993). Interestingly, some fundamental intellectual factors can be found throughout different psychometric factor analytic studies. Carroll (1993) analyzed 461 factor-analytic studies presented in the literature up to date. He observed that some factors tend to appear with a significant frequency across different factorial studies. This is observed in different cognitive areas: reasoning, language, memory, visual perception abilities, etc. Table IV presents a summary of these relatively constant factors found across different factor analytic studies. From a neuropsychological perspective, these factors are expected to be impaired in cases of focal brain pathology. Some neuropsychological syndromes are expected to be observed in cases of disruption of these brain systems supporting these basic cognitive factors. Frontal Lobes and Intelligence Long ago it was noted that frontal damage did not result in evident deficits in psychometric intelligence tests (Hebb, 1939; Hebb and Penfield, 1940). This was true even in cases of bilateral frontal lobectomy. It was somehow surprising to find that IQ in patients with frontal lobe damage could be normal (Hebb, 1945). These initial observations carried out during the 1940s have been further documented in neuropsychology (e.g., Brazzelli et al., 1994; Damasio and Anderson, 1993). Milner (1963) reported a mean loss of only 7.2 IQ points following dorsolateral frontal lobectomies, with mean postoperative IQ scores remaining in the average range. This observation meant that either frontal lobes do not have much to do with intelligence or psychometric intelligence tests were not sensitive to frontal lobe deficits ("executive dysfunctions," according to contemporary terminology). Teuber (1972) carried out a rather extensive research project in order to pinpoint the deficits associated with frontal pathology. He compared patients with right frontal,

12 128 Ardila left frontal, and bilateral damage. The results demonstrated that in general, patients with frontal lesions performed as well as other patients in a variety of intelligence tests. Teuber, however, found deficits in some visuoperceptual tests, such as visual search tasks. These defects in visual search have also been pointed out by different authors (e.g., Luria, 1980). Milner (1982) also found impaired performance in same-different comparisons using clicks, flashes, and colors. Further, she pointed out that frontal damage patients have difficulties indicating the recency of an item in a series using either figures or words. Significant difficulties in sequential or temporal memory have been observed in this group of patients, and it has even been proposed that the temporality of behavior represents a core defect in cases of frontal pathology (Fuster, 1989)... In traditional psychometric intelligence tests, performance of frontal damage patients can be normal or near normal. Black (1976) found a mean WAIS verbal IQ of 99.1 and a mean performance IQ of 99.5 in a group of 44 Vietnam veterans who had sustained unilateral frontal lobe shrapnel injuries. Janowsky et al. (1989) described seven patients with various focal frontal lobe lesions who obtained a mean WAIS-R Full Scale IQ of 101. Damasio and Anderson (1993) analyzed 10 patients with frontal lesions (ventrolaeral and dorsolateral) caused by either vascular events or surgical resection for treatment of tumors. The most notable feature of the WAIS-R testing in these patients was the consistent preservation of the cognitive abilities required to perform the various intellectual tasks following frontal lobe damage. By the same token, in normal subjects, low correlations between traditional intelligence test scores and executive function measures have been reported. Welsh et al. (1991) observed in children that most of the executive function tasks (Visual Search, Verbal Fluency, Motor Planning, Tower of Hanoi, Wisconsin Card Sorting Test (WCST), and Matching Familial Figures Test) were uncorrelated with IQ. Visual Search, Verbal Fluency, WCST, and Tower of Hanoi did not correlate with any IQ measure (Verbal, Quantitative, and Nonverbal) from the Iowa Test of Basic Abilities. Using a 300-subject college student sample, Ardila et al. (1998) observed that Verbal Fluency tests presented a low but significant correlation (about 0.20 to 0.25) with some WAIS verbal subtests, particularly Digits, Arithmetic, and Information. However, WCST scores did not correlate at all with the Verbal, Performance, or Full Scale IQ. Ardila etal. (in press) analyzed the correlation between IQ and some executive function measures (WCST, verbal fluency, and Trial Making Test (TMT), Form A and Form B). Fifty 13- to 16-year-old normal children were selected. It was found that verbal fluency tests correlated about 0.30 with Verbal IQ and Full Scale IQ. In the WCST only Persevertive Errors negatively correlated with Verbal IQ and Full Scale IQ. Only two correlations were found to be significant with regard to the TMT: TMT Form B Errors negatively correlated with WISC-R Vocabulary subtest; and TMT A TIme negatively correlated with Performance IQ. Results were interpreted as supporting the assumption that traditional intelligence tests are not fully evaluating executive functions. In general, prefrontal lobe activity has been associated with self-regulation, control of cognition (metacognition), temporal organization of behavior, monitoring of behavior, selective inhibition of responses to immediate stimuli, planning behavior, and control of attention (Brown, 1985; DamasioandAnderson, 1993; Fuster, 1989; Hecaen, 1964; Luria, 1966, 1969, 1973, 1980; Perecman, 1987; Pribram, 1973; Stuss and Benson, 1983, 1986, 1987). These are the abilities not tapped by psychometric intelligence tests. The term executive junction has been proposed to refer to the multi-operational system mediated by prefrontal areas of the brain and their reciprocal cortical and subcortical connecting pathways (Stuss and Benson, 1986). Executive dysfunction may be summarized in two cardinal defects: in controlling behavior and in organizing cognition. Evidently, traditional intelligence tests do not appropriately evaluate executive function disturbances. It has to be concluded that either executive functions should not be included as elements of "intelligent behavior" or psychometric intelligence tests are insufficient in testing for intelligence. They are not sensitive to the most important elements of "intelligence": "To act purposefully" (i.e., controlling and planning behavior) and "to think rationally" (i.e., organizing and directing cognition) according to Wechsler's (1944) definition of intelligence. The conclusion is evident: psychometric intelligence tests do not appropriately appraise intelligence. Or at least, they are not appraising those abilities that from a neuropsychological perspective (and also from the point of view of the Wechsler's intelligence testing) should be understood as the most important elements in intelligence. Memory and Visuoperceptual Abilities One of most basic functions of the cerebral cortex is to encode and store new information (memory). One basic area in assessing cognitive activity refers to memory evaluation (Lezak, 1995; Spreen and Strauss, 1991). The WIS does not appropriately measure memory. Wechsler himself realized the significant shortcoming of his intelligence scale and created a parallel scale directed specifically to measure memory (Wechsler Memory Scale; Wechsler, 1 I 1

13 Intelligence and Neuropsychology 1945). However, the WIS has remained within the domain of psychometric intellectual measures, whereas the Wechsler Memory Scale has been widely used in the neuropsychology domain. It could be argued that some WIS subtests are indeed evaluating memory: the Information subtest measures remote memory and Digits measures immediate memory. It is easy to agree with this perspective. However, there are at least two significant shortcomings in WIS memory appraisal. (1) The memory process is not evaluated (i.e., the ability to encode, store, and retrieve new information). This is the most critical type of memory test in neuropsychology (e.g., Lezak, 1995; Luria, 1976b). (2) At least verbal and nonverbal memory testing should be included ina cognitive evaluation. Evidently, memory assessment using the WIS is insufficient, even though WAIS-lli has attempted, at least partially, to overcome this significant shortcoming. By the same token, WAIS subtests fail to appropriately measure spatial and visuoperceptual abilities. Indeed, some WIS subtests partially tap into spatial and visuoperceptual abilities (e.g., Picture completion). But it is difficult to accept that WIS subtests are good enough to evaluate spatial and visuoperceptual abilities. There are many better tests in the area (see Lezak, 1995). SOME CONTEMPORARY VIEWS ABOUT INTELLIGENCE Even though the issue of intelligence has been a "hot" topic for many years, the main questions remain unsettled. Recently, some integrative papers have been published, attempting to distinguish what is really known and what still remains controversial with regard to intelligence. Neisser et al. 's (1996) paper "Intelligence: Knowns and Unknowns" published in AmeriCan Psychologist represents perhaps the most authoritative report. The paper was prepared by a task force specifically appointed by the Board of Scientific Affairs of the American Psychological Association. Some of the main conclusions presented by Neisser and his 10 expert co-authors are (1) There are many ways to be intelligent, and there are also many conceptualizations of intelligence. (2) Psychometry has been able to measure a wide range of abilities that are distinct from one another and yet intercorrelated. It is possible to describe the complex relationships between these abilities in many different ways. Some authors have searched for a "general intelligence" (g) factor, whereas others have preferred to refer to a set of independent factors. Still others have opted for a hierarchy of factors. (3) Intelligence is correlated with school achievement at a level of about (some 25% of the variance). (4) Like every trait, intelligence is the joint product of genetic and environmental factors. (5) School affects intelligence in many different ways: transmitting specific information, developing certain skills and attitudes. Failure to attend school has negative consequences in intelligence testing. (6) Some biological conditions have clear negative consequences on intelligence. Examples are perinatal complications, exposure to environmental lead, and exposure to high blood levels of alcohol. (7) There is a steady rise across time in intelligence test scores known as the "Flynn effect" (Flynn, 1984, 1987). Mean IQ scores have increased more than 15 points in the last 50 years. Some reasons may be improved nutrition, cultural changes, experience with testing, shifts in schooling or child-rearing practices, or some other unknown factors. (8) Ethnic differences in intelligence reflect complex patterns. No overall generalization about them is appropriate. (9) Many of the most critical questions about intelligence remain unanswered. Some brief comments may be presented to these selected conclusions: (1) Evidently, the concept and interpretation of intelligence continue to be controversial. Neisser et al. (1996) recognized that there are different ways to interpret intelligence. No single interpretation of intelligence testing data is widely accepted. (2) Neisser et al. (1996) refer to the bidirectional relationship between school and IQ: intelligence predicts school achievement, and school affects intelligence. (3) Many factors may be simultaneously acting on the scores obtained in intelligence tests: genetic factors, some early biological conditions, environmental factors, cultural values, etc. (4) Given the socalled Flynn effect, several factors may be simultaneously interacting to cause the recent rapid rise in test scores. For a person coming from a nonpsychometrically oriented culture (as the author of this paper), however, it is evident that the most crucial factor may be the tremendous training in testing abilities that Americans have been progressively exposed to. Although exposure to psychometric testing has extended to other countries, it is markedly higher in the United States than in most countries. School children currently spend a significant amount of time developing those strategies required in answering tests, and in practicing tasks similar to those included in intelligence tests. This was not observed one generation ago. (5) No generalization or general conclusion about ethnic differences in intelligence is acceptable. There are many ways to be intelligent in different cultural contexts. Good performance on psychometric intelligence tests is just one way to be intelligent in a quite specific cultural context. Reactions to Neisser et al.'s (1996) paper rapidly appeared (see American Psychologist, 52(1), 1997). Reactions were so mixed that the only conclusion that can be

14 130 Ardila safely drawn is that intelligence continues as a very controversial and, in many regards, a poorly understood topic. If a conservative paper like that of Neisser et al. could trigger so many different and opposite reactions, it must be concluded that the concept of intelligence is on very fragile ground. WHAT SHOULD BE INCLUDED WHEN TESTING FOR COGNITIVE ABILITIES? Since Thurnstone (1938, 1947), there is the converging consensus that some fundamental cognitive abilities may be distinguished. Researchers refer to a limited number of domains, usually six to nine, frequently appearing in f.actor analytic studies of psychological (Carroll, 1993) and neuropsychological test batteries (Ardila et al., 1994, 1998; Ponton et al., 1994). A similar idea is presented by Gardner (1983) when he proposed different types of intelligence. Evidently, these are the cognitive domains that should be included when testing for intellectual abilities. There are no fixed tests to evaluated these domains, even though some tests may be better, at least at a certain historical moment. In the future, new and better tests can be developed to appraise these domains, and these domains may even be restated and rearranged. In neuropsychology there are several tests that have become widely accepted and extensively used (see Lezak, 1995; Spreen and Strauss, 1998). They are considered reliable, sensitive, and in general "good" tests. There is a significant research body supporting their reliability and validity. An evaluation of cognitive abilities should include these widely accepted tests. As a matter of fact, many of them have been taken from the intelligence testing research, and in this regard, psychometric intelligence testing and neuropsychological testing may be complementary rather than mutually exclusive. Of course, it is expected that in the future, superior testing instruments will be developed, replacing the current tests that now are considered the best available neuropsychological instruments. Examples of these cognitive domains, and potentially useful tests are: 1. Attention 1.1. Focused attention (e.g., digits backwards) 1.2. Sustained attention (e.g., serial subtractions etc.) 2. Language 2.1. Verbal fluency (using semantic and phonological categories) 2.2. Language comprehension (token test) 2.3. Lexical knowledge (naming, vocabulary, or other similar tests) 3. Calculation abilities 3.1. Arithmetical operations 3.2. Numerical problems 4. Perceptual abilities 4.1. Visual recognition of figures under different conditions (e.g., visual detection, to recognize embedded or unusually presented figures, to find similarities and differences between figures, etc.) 4.2. Recognition of sounds and music (verbal-phonological discrimination; and nonverbal rhythms, melodies, music, etc.) 5. Memory and learning 5.1. Verbal learning (Serial Verbal Learning, California Verbal Learning Test, Rey Auditory Verbal Learning test, Logical Memory, etc.) 5.2. Nonverbal learning (Benton Visual Retention test, immediate and delayed recall of figures) 6. Visuoconstructive and visuospatial abilities 6.1. Visuoconstructive (such as Rey-Osterrieth Complex Figure) 6.2. Tests for spatial abilities (such as line orientation) 7. Motor 7.1. Fine movements (such as the Finger Tapping Test or other fine movements test) 7.2. Praxis ability tests 8. Executive function abilities 8.1. Abstraction (e.g., Similarities) 8.2. Reasoning (e.g., Raven Progressive Matrixes) 8.3. Concept formation tests (the Category Test, Wisconsin Card Sorting Test, etc.) 8.4. Some tests directed to "maintain instructions" (Stroop test, Trial MakingTestFormB, Luria's opposite reactions, etc.) Of course, these tests are not necessarily evaluating a single cognitive domain. Attention is required for an appropriate performance in any intellectual test. Calculation abilities represent a rather complex and multifactorial ability. Verbal memory depends on language understanding. For example, phonological verbal fluency can be interpreted as an executive function test, whereas semantic verbal fluency is closer to a lexical knowledge test. Furthermore, all these tests are significantly influenced by education, age, and cultural background. Norms for different groups should be obtained. Although raw scores can be nonequivalent in different educational, cultural,

15 Intelligence and Neuropsychology and age groups, standard normalized scores are equivalent. Each group itself represents its own norm. Tests must be standardized and norms obtained not only for different age ranges but also for different educational and cultural groups. Otherwise, what is normal for one group might be interpreted as pathological for another. When a particular group outscores another, this simply means that wrong norms have been used. THEWAIS-ill In the last 50 years, the same 11 subtests proposed by Wechsler in the 1930s were repeated over and over again, with just minor changes. Not until 1997 were some fundamental changes introduced to the WIS. The WAIS-III (Wechsler, 1997) represents a very significant improvement over previous versions. But changes are far from enough. The WAIS-III includes some new subtests directed to overcome some of the limitations found in the previous WIS versions: Instead of 11 subtests, there are 14 (seven verbal subtests and seven performance subtests) used to calculate IQs. In the Verbal Scale, the Letter-Number Sequencing subtest (combinations of numbers and letters are read and the examinee is required to recall the numbers first in ascending order and then letters in alphabetical or- 131 der) is added. In the Performance Scale Matrix, Reasoning (the examinee looks at a matrix from which a section is missing and either identifies the number or points to one of five response options that completes the matrix) and Symbol Search (the examinee visually scans two groups of symbols, a target group composed of two symbols and a search group composed by five symbols, and indicates whether either target symbol matches any of the symbols in the search group) were added. The Digit-symbol subtest is used under different conditions: Digit symbol-coding (as previously used), Digit symbol-incidental Learning (paring to recall the symbols matched with the numbers, and free recall to recall the symbols used in coding section), and Digit symbol-copy (to copy the symbols that were used in the Digit symbol-coding). These two last conditions are optional and are not used in calculating IQs. Table V presents the cognitive domains and areas included in the WAIS-III subtests. Several major concerns arise: (1) Testing for cognitive abilities is still insufficient (see above); (2) It is not easy to understand why if several factor analyses disclosed four different factors (Verbal Comprehension, Perceptual Organization, Working Memory, and Processing Speed), the WAIS-III insists on calculating only two compound scores (Verbal and Performance) instead of four. In the different factor analytic studies that are presented in the WAIS-III Technical Manual, the Object Assembly subtest Table V. Cognitive Domains and Areas Included in the WAIS-llI Subtests. Cognitive Domain Area WAIS-llI Subtest I. Attention Focused attention Sustained attention 2. Language Fluency Language comprehension Lexical knowledge 3. Calculation abilities Arithmetical operations Numerical problems 4. Perceptual abilities Visual recognition Auditory recognition 5. Memory Verbal learning Nonverbal learning 6. Visuoconstructive Visuoconstructive and visuospatial Spatial abilities 7. Motor Fine movements Praxis 8. Executive functions Abstraction Reasoning Concept formation "Maintain instructions" Digit span Digit-symbol Not included Not included Vocabulary Not included Arithmetics Picture completion Symbolsearcb Not included Not included Digit-symbol-IT.. Block Design Object Assembly Not included Not included Not included Similarities Matrix reasoning Not included Letter-number Working Memory Processing Speed Verbal Comprehension Working Memory Perceptual Organization Processing Speed Perceptual Organization Verbal Comprehension b Perceptual Organization Working Memory DFactors found in the WAIS-III (Wechsler, 1997). bcomprehension is a subtest difficult to interpret and includes two different types of subtests (proverb interpretation and knowledge of social conventions).

16 132 was not included. No reason for this exclusion is mentioned. (3) Scores are corrected according to age but not according to educational level, although the most important variable affecting psychological and neuropsychological test performance is education, not age (e.g., Anastasi, 1988; Cronbach, 1990; Ostrosky et al., 1998). Nevertheless, raw scores should be corrected by both education and age. In the standardization sample used, about two third,s of the subjects had 12 or more years of education. The lowest educational group included in the normative sample had "eight or less years of education," which is most likely an inappropriate education cut-off point (Ostrosky et al., 1998). The WAIS-ID seems inappropriate to test people with low educational levels. (4) The WAIS-ID retains the total compound score (Full Scale IQ) despite tj;te fact that the factor analytic studies presented in the WAIS-ID Technical Manual show that the four factors obtained in the different factor analyses are rather independent. In brief, the WAIS-ID is undoubtedly an improvement over previous WIS versions. This is the first time that Wechsler's original testing schema has been, at least partially, abandoned, and new tests are included. After almost 50 years of extensive use, it was finally accepted that the 11 WIS subtests proposed by Wechsler were insufficient to test cognitive abilities. Evidently, WAIS-ID represents an implicit recognition that at least executive functions and memory had been insufficiently tested in the WIS previous versions. Nonetheless, it does not mean that new WAIS-ID has overcome the major difficulties pointed out above, and no research studies using the WAIS-ID are yet available. COGNITIVE ABILITIES IN DIFFERENT CULTURAL CONTEXTS Psychometric intelligence tests have been developed in a very specific cultural context. They rely on those abilities that are significant in that particular cultural context and on those approaches that are culturally most valuable. In neuropsychology, cognitive disturbances associated with brain pathology. of a very limited subsample of the human species-contemporary Western, and most often, urban middle-class, and literate brain-damaged individuals-have been relatively well analyzed. Our understanding about the brain's organization of cognitive abilities, and their disturbances in cases of brain pathology, is therefore not only partial but, undoubtedly, culturally biased. Cultural and linguistic diversity is an enormous, but frequently overlooked, moderating variable. Several thousand different cultures have been described by Ardila anthropology (e.g., Bernatzik, 1957), and contemporary humans speak over 4,000 different languages (Swadesh, 1967). Evidently, an extended analysis of cognitive disturbances in different cultural and ecological contexts is necessary for us to understand and serve the neuropsychological needs of our constituency. Supposedly, comparable fundamental cognitive disturbances are to be found in every human species member, regardless of cultural background, educational level, language, and ecological demands, as a consequence of brain lesions. There are some fundamental characteristics in the human brain, and in brain-behavior relationships that one would expect to observe in every human subject. Basic cognitive processes are universal, and cultural differences in cognition reside more in the situations to which particular cognitive processes are applied than in the existence of the process in one cultural group and the absence in another. Culture prescribes what should be learned and at what age. Consequently, different cultural environments lead to the development of different patterns of abilities. Cultural and ecological factors play a role in developing different cognitive styles (Berry, 1971, 1979). Furthermore, cultural variables can eventually influence the brains' organization of cognition. For example, it has been reported that the degree (not the direction) of brain lateralization of language can depend on literacy, and in general, on the verbal training histories (Lecours et al., 1987, 1988; Matute, 1988). Language, memory, visuospatial abilities, praxis skills, and cognitive abilities in general studied in psychology and neuropsychology are under cultural influence. Cognitive disturbances associated with brain pathology are related to the way those abilities have been trained. Brain damage results in the disturbance in a specific level ofinfonnation processing, even though the actual manifestation depends on the specific pattern of cognitive abilities. This specific pattern of cognitive abilities is culturally dependent. Reading and writing can illustrate the enonnous complexity of brain organization of any cognitive process. Supposedly, reading is based on certain fundamental abilities (e.g., complex shape perception, cross-modal learnings, etc.) already existing 5,000 years ago in preliterate humans, and of course, existing in illiterate individuals. The human brain might be specialized not for reading or writing per se, but for certain basic abilities (infonnation processing levels or cognition factors) required to read and to write, albeit not only to read or to write. It does not seem reasonable to assume that some brain areas are specialized in reading or writing, just as it does not seem reasonable to suppose that some brain areas are specialized in using computers. Even though the fundamental defect may be the same, the actual manifestation of brain pathology may

17 Intelligence and Neuropsychology be different depending on the training histories and the actual pattern of cognitive abilities. If, despite some existing basic characteristics in its brain organization, oral and written language disturbances are associated with language idiosyncracies (e.g., aphasia is not completely equivalent in Chinese and Spanish; alexia and agraphia can be different in English, Spanish; and Japanese, etc.; Ardila et al., 1996; Sasanuma and Fujimura, 1971; Yamadori, 1975; Yu-Huan et al., 1990), other cognitive abilities such as spatial cognition may also depend on the specific ability learning history and the particular cognitive style. Supposedly, similar spatial cognition disturbances are to be found in a similar way in every human species member, regardless of the cultural batkground and the ecological demands. Basic spatial cognition abilities, however, are applied in rather different ways depending on the specific cultural context and the ecological demands. Contemporary city individual's spatial abilities are not necessarily inferior (or superior) to the Eskimos' or Amazonian Indians' spatial abilities. Spatial abilities may have evolved with new living and cultural conditions. Spatial cognition abilities can be required in many contemporary, and historically recent skills, not found in every human group. The author of this paper had the opportunity to study a university chemistry professor who suffered a small right-parietal infarction without any overt spatial disturbance. Although she had not any evident spatial difficulty in her everyday activities, she could not continue teaching chemistry, because she was "unable to have a spatial representation of molecules and all the time got confused." Mathematics (Ardila and Rosselli, 1990; Luria, 1980), painting, chemistry, chess (Chabris and Hamilton, 1992), reading and writing (Ardila and Rosselli, 1993; Benson and Ardila, 1996), mechanics (Benton, 1989), and even music (Henson, 1985) all represent, at least partially, spatially based skills. They can be impaired in cases of right hemisphere damage of those same areas that in an Eskimo or Amazonian Indian would imply an impossibility to move around the snow or the jungle. The analysis of cognitive disturbances associated with brain pathology in different cultural contexts, no doubt, represents a challenge for 21st century neuropsychology. This analysis may eventually allow researchers to have a better understanding of the basic elements (factors) of cognition. SOME TENTATIVE CONCLUSIONS 1. Psychometric intelligence tests (e.g., WIS in its different versions) do not seem to measure what from 133 a neuropsychological perspective (and also according to Wechsler himself) could be better interpreted as "intelligence." At the least, they fail in appraising some most fundamental aspects. 2. The concept of IQ might disappear. It is archaic, and theoretically remains a controversial concept. Subtests used to measure "intelligence" are inappropriate. 3. In the future, cognitive evaluation may rely on neuropsychological instruments instead of using psychometric intelligence tests. Neuropsychological tests have a clear and overt rationale from the point of view of the brain organization of cognitive activity. No clear rational for the WIS subtests is easily found. 4. It would seem more appropriate to use standard scores (such as T, z, or percentiles) for the individual tests and cognitive domains, than using global intellectual scores (such as IQ). ACKNOWLEDGMENTS My most sincere gratitude to VIrginia Standish for her most valuable support and help while preparing this paper. Thanks to Dr. Kevin Keating, Dr. Mariano Alemagny, and Dr. Sarah Ransdell for their observations and suggestions on this paper. My gratitude to the anonymous reviewers for their most helpful, encouraging, and interesting observations. REFERENCES Anastasi. A. (1988). Psychological Testing (6th ed.). MacMillan. New York. Ardila, A. (1995a). Directions of research in cross-cultural neuropsychology. Journal of Clinical and Experimental Neuropsychology 17: Ardila, A. (1995b). Estructura de la actividad cognoscitiva: Hacia una teoria neuropsicol6gica [Structure of cognitive activity: Toward a neuropsychological theory]. Neuropsychologia Latina 1(2): Ardila, A. Galeano. L. M. and Rosselli. M. (1998). Toward a model of neuropsychological activity. Neuropsychology Review 8: Ardila, A. Pineda, D. and Rosselli. M. (in press). Correlation between intelligence test scores and executive function measures. Archives of Clinical Neuropsychology. Ardila, A. and Rosselli. M. (1990). Acalculias. Behavioural Neurology 3: Ardila, A. and Rosselli. M. (1993). Spatial agraphia. Brain and Cognition 22: Ardila, A. and Rosselli. M. (1994). Development of language.!pemory and visuospatial abilities in 5- to 12-year-old children using a neuropsychological battery. Developmental Neuropsychology 10: Ardila, A. Rosselli. M. and Bateman. J. R. (1994). Factorial structure of cognitive activity. Behavioral Neurology 7: Ardila, A. Rosselli. M. and Ostrosky. F. (1996). Agraphia in Spanishlanguage. Aphasiology 10:

18 134 ArdiJa Benson, D. F. (1982). The alexia: A guide to the neural basis of reading. In IGrsher, H. S., and Freemon, F. (eds.), The Neurology of Aphasia, Swets, Amsterdam. pp Benson, D. F. (1994). The Neurology of Thinking, Oxford University Press, New York. Benson, D. F., and Ardila, A. (1996). Aphasia: A Clinical Perspective, Oxford University Press, New York. Benton, A. (1989). Constructional apraxia. In Goodglass, H., and Damasio, A. R. (005.), Handbook of Clinical Neuropsychology (Vol. 2), Elsevier, Amsterdam, pp Bematzik, H. A. (1957). Razas y pueblos del mundo [World races and people]. Vols Ediciones Ave, Barcelona. Beny, 1. W. (1971). Ecological and cultural factors in spatial perceptual development. Canadio.n Journal of Behavioral Sciences 3: Beny, 1. W. (1979). Culture and cognition style. In Mesella, A., Tharp, R. G., and Ciborowski, T. 1. (005.), Perspectives in Cross-Cultural Psychology, Academic Press, New York, pp Beny, J. W., Pooninga, Y. H., Segall, M. H., and Dasen, P. R. (1992). Cross-Cultural Psychology, Cambridge University Press,.. Cambridge. Binet, A. (1905). Analysis of C.E. Spearman: The proof and measurement of association between two things and general intelligence objectively determined and measured. L 'Anne Psychologique II: Binet, A. (1908). Le developpement de I'intelligence chez les enfants. [The development of intelligence in children.] L'Anne Psychologique 14: Black, F. W. (1976). Cognitive deficits in patients with unilateral war-related frontal lobe lesions. Journal of Clinical Psychology 32: Bomstein, R. A., and Chelune, G. 1. (1988). Factor structure of the Wechsler Memory Scale-Revised. The Clinical Neuropsychologist 2: Brazzelli, M., Colombo, N., Della SaIIa, S., and Spinnler, H. (1994). Spared and impaired abilities after bilateral frontal damage. Cortex 30: Brody, N. (1992). Intelligence (2nd ed.), Academic Press, New York. Brown, J. E. (1985). Frontal lobe syndromes. In Frederiks J. A. M. (ed.), Handbook of Clinical Neurology, Vol. 45: Clinical Neuropsychology, Elsevier, Amsterdam, pp Carroll, J. B. (1993). Human Cognitive Abilities: A Survey of Factor Analytic Studies, Cambridge University Press, Cambridge. Cattell, R. B. (1971). Abilities: Their Structure, Growth and Action, Houghton-Mifflin, Boston. Ceci, S. J. (1990). On Intelligence.. More or Less: A Bioecological Treatise on Intellectual Development, Prentice Hall, Englewood, NJ. Ceci, S. J. (1991). How much does schooling influence general intelligence and its cognitive components? A reassessment of evidence. Developmental Psychology 27: Ceci, S. J., and Williams, W. M. (1997). Schooling, intelligence and income. American Psychologist 52: Chabris, C. F., and Hamilton, S. E. (1992). Hemispheric specialization for skilled perceptual organization by chessmasters. Neuropsychologia 30: Cronbach, L. J. (1990). Essentials of Psychological Testing (5th ed.), Harper and Row, New York. Damasio, A. R., and Anderson, S.W. (1993). The frontal lobes. In Heilman, K. M., and Valenstein, E. (eds.), Clinical Neuropsychology, Oxford University Press, New York, pp rd edition. Damasio, H., and Damasio, A. R. (1989). Lesion Analysis in Neuropsychology, Oxford University Press, New York. Das, J. P., Naglieri, J. A., and Kirby, J. R. (1994). Assessment of Cognitive Processes: The PASS Theory of Intelligence, Allyn & Bacon, Needham Heights, MA. De Renzi, E. (1982). Disorders of Space Exploration and Cognition, Wiley, New York. De Renzi, E. (1985). Disorder of space exploration. In Frederiks, J. A. M. (ed.), Handbook of Clinical Neurology, Vol. 45: Clinical Neuropsychology, Elsevier, Amsterdam, pp Elwood, R. W. (1991). Factor structure of the Wechsler Memory Scale-Revised (WMS-R) in a clinical sample: A methodological reappraisal. The Clinical Neuropsychologist 5: Eysenck, H. (1986). Inspection time and intelligence: A historical introduction. Personality and Individual Differences 7: 603-'607. Flynn, J. R. (1984). The mean IQ gains of Americans: Massive gains Psychological Bulletin 95: Flynn, 1. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin 101: Frearson, W. M., and Eysenck, H. J. (1986). Intelligence, reaction time [TR] and a new "odd-man-out" RT paradigm. Personality and Individual Differences 7: Fuster, 1. M. (1989). The Prefrontal Cortex (2nd ed.), Raven Press, New York. Gardner, H. (1983). Frames of Mind: The Theory of Multiple Intelligences, Basic Books, New York. Greenfield, P. M. (1997). You can't take it with you: Why ability assessments don't cross cultures. American Psychologist 52: Guilford, 1. P. (1967). The Nature of Human Intelligence, McGraw-HilI, New York. Guilford, 1. P. (1968). Intelligence has three facets. Science 160: Guilford, J. P., and Hoepfner, R. (1971). The Analysis of Intelligence. McGraw-HilI, New York. Hebb, D. (1939). Intelligence in man after large removals of cerebral tissue: Report of four left frontal lobe cases. Journal of General Psychology 21: 73-f.7 Hebb, D. O. (1942). The effects of early and late brain injury upon test scores and the nature of normal adult intelligence. Proceedings of the American Phylosophical Society 85: Hebb, D. O. (1945). Man's frontal lobe: A critical review. Archives of Neurology and Psychiatry 54: Hebb, D., and Penfield, W. (1940). Human behavior after extensive bilateral removal from the frontal lobes. Archives of Neurology and Psychiatry 44: H~aen, H. (1962). Clinical symptomatology in right and left hemisphere lesions. In Mountcastle, V. B. (ed.), Interhemispheric Relations and Cerebral Dominance, John Hopkins University, Baltimore, pp Hecaen, H. (1964). Mental symptoms associated with tumors of the frontal lobe. In Warren, 1. M., and Akert, K. (eds.), The Frontal Granular Cortex and Behavior, McGraw-Hili, New York, pp H~aen, H., and Albert, M. L. (1978). Human Neuropsychology; Wiley, New York. Henson, R. A. (1985). Amusia. In Frederiks, J. A. M. (ed.), Handbook of Clinical Neurology, Vol. 45: Clinical Neuropsychology, Elsevier, Amsterdam, pp Hunter, J. (1986). Cognitive ability, cognitive aptitudes, job knowledge, and job performance. Journal of Vocational Behavior 29: Irvine, S. H., and Berry, J. W. (eds.), (1988). Human Ability in Cultural Context, Cambridge University Press, Cambridge. Janowsky, J. S, Shimamura, A. P., Kritchevsky, M., and Squire, L. R. (1989). Cognitive impairments following frontal lobe damage and its relevance to human amnesia. Behavioral Neurosciences 103: Jensen, A. R. (1980). Bias in mental testing, Free Press, New York. Jensen, A. R. (1987). Individual differences and the Hick paradigm. In P. A. Vernon (ed.), Speed of Information Processing and Intelligence, Ablex, Norwood, pp Kaiser, H. F. (1960). The varimax solution for Primary Mental Abilities. Psychometrika 25: Kaplan, E., Fein, D., Morris, R., and Delis, D. (1991). WAlS-R as a Neuropsychological Instrument, Psychological Corporation, San Antonio, TX.

19 InteUige~ce and Neuropsychology 135 Kaufman, A. S., and Kaufman, N. L. (1993). Manual for the Kaufman Adolescent andadult Intelligence Test (KAlTJ, American Guidance Service, Circle Pines, MN. Kaufman, A. S., and Kaufman, N. L. (1997). The Kaufman Adolescent and Adult Intelligence TesL in Flanagan, D. P., GensJtafi, J. D., and Harrison, P. L. (eds.), Contemporary Intellectual Assessment: Theories, Tests and Issues, Guilford Press, New York, pp Kertesz, A. (1983). Localization in Neuropsychology, Academic Press, New York. Lecours, A. R., Mehler, 1., Parente, M. A., Beltrami, M. C., Canossa de Tolipan, L., Castro M. J., Carrano, V., Chagastelles, L., Dehaut, F., Delgado, R., Evangelista, A., Fajgenbaum, S., Fontoura, C., de Fraga Karmann, D., Gurd, J., Hierro Tome, C., Jakubovicz, R., Kac, R., Lefevre, B., Lima. C., Maciel, J., Mansur, L., Martinez, R., Nobrega, M. C., Osorio, Z., Paciomik, J., Papaterra, F., Jourdan Penedo, M. A., Saboya, B., Scheuer, C., Batista da Silva, A.,Spinardi, M., and Texeira, M. (1988). llliteracy and brain damage, m: A contribution to the study of speech and language disorders in illiterates with unilateral brain damage (initial testing)..; Neuropsychologia 26: Lecours, R. L., Mehler, J., Parente, M. A., Caldeira, A., Cary, L., Castro, M. J., Dehaout, F., Delgado, R., Gurd, J., Karmann, D., Jakubovitz, R., Osorio, z., Cabral, L. S., and Junqueira, M. (1987). llliteracy and brain damage, I: Aphasia testing in culturally contrasted populations (control SUbjects). Neuropsychologia 25: Lezak, M. D. (1995). Neuropsychological Assessment (2nd ed.), Oxford University Press, New York. Luria, A. R. (1966). Human Brain and Psychological Processes, Harper and Row, New York. Luria, A. R. (1969). Frontal lobe syndromes. In Vmken, P. J., and Bruyn, G. W. (eds.), Handbook of Clinical Neurology, Vol. 2, Nonh Holland, Amsterdam, pp Luria, A. R. (1973). The frontal lobes and the regulation of behavior. In Pribrarn, K. H., and Luria, A. R. (eds.), Psychophysiology of the Frontal Lobes, Academic Press, New York, pp Luria, A. R. (l976a). Basic Problems ofneurolinguistics, Mouton, The Hague, Netherlands. Luria, A. R. (I 976b). The Neuropsychology of Memory, Halstead, Somerset. NJ. Luria, A. R. (1980). Higher Cortical Functions in Man (2nd ed.), Basic. New York. Matarazzo. 1. D. (1972). Wechsler's Measurement and Appraisal of Adult Intelligence, Oxford University Press. New York. Matute. E. (1988). EI aprendizaje de la lectoescritura y la especializaci6n hemisf6rica para el lenguaje [Reading and writing learning, and hemispheric specialization for language]. In Ardila, A., and Ostrosky-Solis. F. (eds.). Lenguaje oral y escrito [Oral and wrinen language]. Mexico D. F., Mexico: Editorial Trillas. pp Milner, B. (1963). Effects of different brain lesions on card sorting. Archives of Neurology 9: Milner, B. (1982). Some cognitive effects of frontal-lobe lesions in man. Philosophical Transactions of the Royal Society of London 298: Morrow, L., and Ratcliff, G. (1988). The neuropsychology of spatial cognition. In Stiles-Davis, 1., Kritchevsky, M., and Bellugi, U. (eds.), Spatial Cognition: Brain Bases and Development, Erlbaum, Hillsdale, Nl, pp Naglieri, 1. A. (1997). Planning, attention, simultaneous and succesive theory and the cognitive assessment system: A new theory-based measure of intelligence. In Flanagan, D. P., Genshaft, 1. D., and Harrison, P. L. (eds.), Contemporary Intellectual Assessment: Theories. Tests and Issues, Guilford Press, New York, pp Naglieri, 1. A., and Das, 1. P. (1996). Das Naglieri Cognitive Assessment System, Riverside, Chicago. Neisser, U., Boodoo, G., Bouchard, T. 1., Boykin, A. W., Brody, N., Ceci, S. 1., Halpern, D. E, Loehlin, 1. C., Perloff, R., Stenberg, R. 1., and Urbina, S. (1996). Intelligence: Knowns and unknowns. American Psychologist 51: Nett1ebeck, T. (1987). Inspection time and intelligence. In Vernon, P. A. (ed.), Speed of Information Processing and Intelltgence, Ablex, Norwood, NJ, pp Newcombe, F., and Ratcliff, G. (1989). Disorders ofvisuospatial analysis. In Goodglass, H., and Damasio, A. R. (eds.), Handbook of CUnical Neuropsychology (Vol. 2), Elsevier, Amsterdam, pp Ostrosky, F., Ardila, A., and Rosselli, M. (1998). Neuropsychological test performance in illiterates. Archives of CUnical Neuropsychology 13: 645~. Ostrosky, F., Ardila, A., and Rosselli, M. (1999). NEUROPSI: A brief neuropsychological test battery in Spanish with norms by age and educational level. Journal of the International Neuropsychological Society 5: Perecman, E. (1987). Consciousness and the meta-functions of the frontal lobes. In Perecman, E. (ed.), The Frontal Lobes Revisited, IRBN Press, New York, pp Pettersen, S. E., Fox, P. T., Snyder, A. Z., and Raichle, M. E. (1989). Activation of extrastriate and frontal cortical areas by visual words and word-like stimuli. Science 249: Pirozzolo, F. J. (1985). Mental retardation. In Frederiks, 1. A. M. (ed.) Handbook of Clinical Neurology, Vol. 46: Neurobehavioral Disorders, Elsevier, Amsterdam, pp Pont6n, M. 0., Satz, P., and Herrera, L. (1994). Factor analysis of a neuropsychological screening battery for Hispanics (NeSBHIS). XVII European Meeting of the International Neuropsychological Society, June. Posner, M. I., Petersen, S. E., Fox, P. T., and Raichle, M. E. (1988). Localization of cognitive operations in the human brain. Science 240: Pribram, K. H. (1973). The primate frontal cortex-executive of the brain. In Pribram, K. H., and Luria, A. R. (eds). Psychophysiology of the Frontal Lobes, Academic, New York, pp Price, C. 1., Wise, R. 1. S., Watson, 1. D. G., Patterson, K., Howard, D., and Frackowiak, R. S. 1. (1994). Brain activation during reading: The effects of exposure duration and task. Brain 117: Reed, T. E., and Jensen, A. R. (1992). Conduction velocity in a brain nerve pathway of normal adults correlates with intelligence level. InteUigence 16: Roid, G. H., Prifitera, A., and Ledbetter, M. (1988). Confirmatory analysis of the factor structure of the Wechsler Memory Scale-Revised. Clinical Neuropsychology 2: Rosselli. M., and Ardila, A. (1989). Calculation deficits in patients with right and left hemisphere damage. Neuropsychologia 27: Sasanuma, S., and Fujimura, O. (1971). Kanji versus Kana processing in alexia with transient agraphia: A case report. Conex 7: Spearman, C. (1904). "General intelligence" objectively determined and measured. American Journal of Psychology 15: Spearman, C. (1923). The Nature of "Intelligence" and the Principles of Cognition, MacMillan, London. Spreen, 0., and Strauss, E. (1991). A Compendium of Neuropsychological Tests, Oxford University Press, New York. Spreen, 0., and Strauss, E. (1998). A Compendium of Neuropsychological Tests: Administration, Norms and Commentary (2nd ed.), Oxford University Press, New York. Stem, W. (1912). Die psychologischen Methoden der Intelligenz prufungo [The Psychological Method to Measure Intelligence.] Leipzig: Barth. Sternberg, R. 1. (1985). Beyond IQ: A Triarchic Theory of Human Intelligence, Cambridge University Press, New York. Sternberg, R. 1. (1988). A triarchic view of intelligence. In Irvine, S. H., and Berry, 1. W. (eds.), Human Ability in Cultural Context, Cambridge University Press, New York, pp Sternberg, R. J. (1997). The concept of intelligence and its role in lifelong learning and success. American Psychologist 52: Stuss, K. H., and Benson, D. F. (1983). Frontal lobe lesions and behavior. In Kertesz, A. (ed.), Localization in Neuropsychology, Academic Press, New York, pp z